SpatialLadder: Progressive Training for Spatial Reasoning in Vision-Language Models

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Li, Hongxing, Li, Dingming, Wang, Zixuan, Yan, Yuchen, Wu, Hang, Zhang, Wenqi, Shen, Yongliang, Lu, Weiming, Xiao, Jun, Zhuang, Yueting
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866908585252880384
author Li, Hongxing
Li, Dingming
Wang, Zixuan
Yan, Yuchen
Wu, Hang
Zhang, Wenqi
Shen, Yongliang
Lu, Weiming
Xiao, Jun
Zhuang, Yueting
author_facet Li, Hongxing
Li, Dingming
Wang, Zixuan
Yan, Yuchen
Wu, Hang
Zhang, Wenqi
Shen, Yongliang
Lu, Weiming
Xiao, Jun
Zhuang, Yueting
contents Spatial reasoning remains a fundamental challenge for Vision-Language Models (VLMs), with current approaches struggling to achieve robust performance despite recent advances. We identify that this limitation stems from a critical gap: existing methods attempt to learn spatial reasoning directly without establishing the hierarchical foundations of perception and understanding. To address this challenge, we present a comprehensive methodology for building spatial intelligence progressively. We introduce SpatialLadder-26k, a multimodal dataset containing 26,610 samples spanning object localization, single image, multi-view, and video spatial reasoning tasks, constructed through a standardized pipeline that ensures systematic coverage across modalities. Building on this dataset, we design a three-stage progressive training framework that (1) establishes spatial perception through object localization, (2) develops spatial understanding through multi-dimensional spatial tasks, and (3) strengthens complex reasoning via reinforcement learning with verifiable rewards. This approach yields SpatialLadder, a 3B-parameter model that achieves state-of-the-art performance on spatial reasoning benchmarks, with 23.4% average improvement over the base model, surpassing GPT-4o by 20.8% and Gemini-2.0-Flash by 10.1%. Notably, SpatialLadder maintains strong generalization with 7.2% improvement on out-of-domain benchmarks, demonstrating that progressive training from perception to reasoning is essential for robust spatial intelligence.
format Preprint
id arxiv_https___arxiv_org_abs_2510_08531
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle SpatialLadder: Progressive Training for Spatial Reasoning in Vision-Language Models
Li, Hongxing
Li, Dingming
Wang, Zixuan
Yan, Yuchen
Wu, Hang
Zhang, Wenqi
Shen, Yongliang
Lu, Weiming
Xiao, Jun
Zhuang, Yueting
Computer Vision and Pattern Recognition
Artificial Intelligence
Computation and Language
Spatial reasoning remains a fundamental challenge for Vision-Language Models (VLMs), with current approaches struggling to achieve robust performance despite recent advances. We identify that this limitation stems from a critical gap: existing methods attempt to learn spatial reasoning directly without establishing the hierarchical foundations of perception and understanding. To address this challenge, we present a comprehensive methodology for building spatial intelligence progressively. We introduce SpatialLadder-26k, a multimodal dataset containing 26,610 samples spanning object localization, single image, multi-view, and video spatial reasoning tasks, constructed through a standardized pipeline that ensures systematic coverage across modalities. Building on this dataset, we design a three-stage progressive training framework that (1) establishes spatial perception through object localization, (2) develops spatial understanding through multi-dimensional spatial tasks, and (3) strengthens complex reasoning via reinforcement learning with verifiable rewards. This approach yields SpatialLadder, a 3B-parameter model that achieves state-of-the-art performance on spatial reasoning benchmarks, with 23.4% average improvement over the base model, surpassing GPT-4o by 20.8% and Gemini-2.0-Flash by 10.1%. Notably, SpatialLadder maintains strong generalization with 7.2% improvement on out-of-domain benchmarks, demonstrating that progressive training from perception to reasoning is essential for robust spatial intelligence.
title SpatialLadder: Progressive Training for Spatial Reasoning in Vision-Language Models
topic Computer Vision and Pattern Recognition
Artificial Intelligence
Computation and Language
url https://arxiv.org/abs/2510.08531